I'm a researcher at Intel Research Seattle (that means I do research and then convince a company that has lots of money to build great things with that research).
I'm also affiliate professor in the University of Washington Computer Science and Engineering department (that means I can check out books from the library, teach classes, advise students, and use the espresso machine in the AI lounge).
I develop machine learning algorithms that interpret the output of sensors. I use cameras, radio-frequency, and sonar sensors to identify and locate objects in our environment. These sensors are particularly interesting because the relationship between what they report (pixel values, signal strengths, etc) and what we care about (position or identity of objects) is very difficult to model. Rather than relying on detailed (and typically unrealistic) physics-based models, I build systems that learn models from raw data. Building these methods requires tools from dynamical systems, optimal controls, convex optimization, and approximation algorithms. It is a complicated area that makes you grow hair on your chest and makes the hair on your head go white (or in my case, make it disappear altogether).
More details on my research publications.
For more information, see my personal web page
CSE599R course web page.